Anthropic quietly launched its new flagship model Claude Mythos 5, now available via the Anthropic API. This is Anthropic’s third major model released in 2026, following Claude Opus 5 and Claude Sonnet 5, positioned at the highest performance tier.
Pricing and availability
Claude Mythos 5 pricing is $11 per million input tokens,$55 per million output tokens,and Claude Slightly lower than Opus 5 ($15/$75), yet still at a premium price point. Currently available only through the official Anthropic API, not yet on third-party platforms such as AWS Bedrock or Google Vertex AI.
core competencies
- Extremely large context window: Supports up to 1 million tokens of context, ideal for long-document processing, codebase analysis, and similar use cases
- Multimodal capabilities: Supports visual input (image understanding), tool use, and prompt caching
- Strong reasoning capabilities: Scores 84.2 on GPQA (graduate-level question answering) and performs robustly on coding benchmarks
- Response speed: ~39 tokens per second, with first-token latency of ~2.5 seconds
and Claude Family Comparison
Anthropic’s current flagship lineup includes Opus 5 ($15/$75), Mythos 5 ($11/$55), and Sonnet 5 ($3/$15). Priced between Opus 5 and Sonnet 5, Mythos 5 is positioned closer to Opus 5 in performance. The name “Mythos” (myth) likely signals Anthropic’s new exploration into reasoning and deep thinking.
Compare with competing products
In today’s flagship model competitive landscape,Claude Mythos 5 is priced lower than OpenAI GPT-5.3 Chat (price not publicly disclosed but industry estimates suggest it is higher), yet higher than DeepSeek domestic open-source/commercial models such as V4 Pro and Qwen3.8 Max. For developers and enterprise users, Mythos 5 offers a new option situated between “most expensive and most capable” and “value-first.”
Summary: A noteworthy new flagship
Claude The launch of Mythos 5 indicates Anthropic is accelerating its product iteration pace. Claude From Opus 5 to Mythos 5, the interval was less than two months—demonstrating Anthropic’s determination to compete in the flagship model arena. For enterprise users handling complex reasoning tasks or long-document analysis, Mythos 5 is a new option worth evaluating. However, third-party platforms do not yet support it, and ecosystem maturity will require time.
Want more AI model evaluations and comparisons? Visit AI Dash ——Discover the best AI tools.